Acute drug toxicity related to drinking herbal tea in a kidney transplant recipient

Abstract

Calcineurin and mTOR inhibitors are commonly used immunosuppressive agents with narrow therapeutic range. As the drugs are mainly metabolized by the P450 cytochrome system, the interaction between food and herbs are also commonly seen and affect the drug levels. We present a case of a kidney transplant recipient with toxic therapeutic levels of cyclosporine A and sirolimus due to interaction between the immunosuppressive agents and Chinese herbal tea. Ingredients within the herbal tea were reported to have inhibitory effect on cytochrome CYP3A4 in-vitro studies. Transplant recipients should be alert that there may be potent interaction between the immunosuppressive drugs and herbs resulting in adverse effect on allograft function.

• Drug interaction
• drug toxicity
• herbal tea
• kidney transplantation

Introduction

Cyclosporine A and sirolimus have narrow therapeutic range. Being metabolized by the P450 cytochrome system, drug–drug interaction has never been less stressed. Yet, the importance of interaction between food and herbs with the drug is frequently forgotten. Herbs are widely used and have always played an important role in the Chinese population. To many, herbs, being natural food sources, are safe to take. Consumption of traditional herbal tea for “maintenance” of health, or for treatment of different illnesses is part and partial in the Chinese culture. Previous case reports1,2 described the herbal tea lowered the cyclosporine A level. However, there is no report stated herbal tea may raise drug levels of cyclosporine A and sirolimus. Herein, we present a case of a kidney transplant recipient had toxic therapeutic levels of cyclosporine A and sirolimus due to intake of Chinese herbal tea.

Case report

A 64-year-old Chinese gentleman with unknown cause of renal failure was first case reported in 2000. Ultrasound of the kidney revealed a small renal stone in the left kidney and renal biopsy was not performed because of bilateral small size kidneys. After diagnosed to have end stage renal failure, he then underwent peritoneal dialysis for a year and received cadaveric kidney transplantation in 2001. His baseline immunosuppressant regimen included prednisolone, tacrolimus and azathioprine. Six months after the transplant, he was diagnosed to have diabetic mellitus and tacrolimus has been converted to cyclosporine A. One year after the transplantation, he was noted to have elevation of serum creatinine from 1.47 mg/dL to 1.83 mg/dL with low trough level of cyclosporine A (136 µg/L). Further investigation confirmed he had acute cellular rejection which was well responded to a course of pulse methylprednisolone and his serum creatinine was back to baseline around 1.47 mg/dL.

Six years later, his illness was complicated by having carcinoma of urinary bladder and required radical cystectomy and creation of ileal conduit. In view of the history of malignancy, the immunosuppressive regime was changed to prednisolone, cyclosporine A and sirolimus. The trough level of cyclosporine A and sirolimus were maintained at around 60–80 µg/L and 8–12 µg/L, respectively. Other medications included valsartan, diltiazem, famotidine, gliclazide and aspirin for his ischemic heart disease. The patient was being regularly followed-up three-monthly, with his renal function and drug levels monitored in every clinic visit.

Ten years after kidney transplant, he was found to have acute raise in serum creatinine from baseline 1.47 mg/dL to 3.43 mg/dL in pre-clinic follow-up test. The estimated glomerular filtration rate (eGFR) dropped from 47 to 18 mL/min/1.73 m². With the usual dosage of cyclosporine A (Neoral®) 30 mg bd (1 mg/kg/day) and sirolimus (Rapamune®) 1 mg daily, the trough levels of cyclosporine A and sirolimus were detected as 789 µg/L and 45 µg/L, respectively. Other blood tests showed urea nitrogen peaked at 70.9 mg/dL, potassium level at 8.2 mEq/L, sodium 133 mEq/L and bicarbonate 16 mEq/L. The urine test for proteinuria was trace and no obvious red blood cell or white cell.

Clinically, patient did not have any fever except some flu-like symptoms. Systemic examination did not reveal any abnormality. He denied any change in his medications but confessed to have taken around 500 ml of herbal tea for 3 consecutive days and two tablets of ibuprofen prescribed by his family doctor for myalgia few days ago. The remnant of the herbal tea was collected and analysis of the tea showed it contained 2-(4-hydroxyphenyl) ethanol, cimicifuga (CF), isoliquiritigeni (IS), oleanolic acid (OA), ursolic acid and other unknown ingredients.

He was immediately admitted and monitored meticulously. Potassium level came down quickly with diuresis, hence dialysis was not given. The patient’s serum creatinine were slowly improved and returned to baseline within 5 d of admission. With regards to the cyclosporine A and sirolimus, both drugs were withheld and the drug levels were closely monitored (Figure 1). At day 3 of hospitalization, cyclosporine A 30 mg BD were reintroduced when its trough level came down to 160 µg/L whereas sirolimus 1 mg daily were given when the trough level came down to 15 µg/L. Allograft biopsy was not performed in view of good recovery of allograft function. Repeated blood tests showed patient has stable allograft function and good therapeutic drug levels and was discharged from the hospital after 10 d of hospitalization. Upon the latest out-patient visit, his allograft function remained stable and the cyclosporine A and sirolimus level were at 79 µg/L and 8.2 µg/L, respectively.

Figure 1. (a) Serial trough level of cyclosporine A and (b) serial trough level of sirolimus.

Discussion

Maintaining the therapeutic drug levels of cyclosporine A and sirolimus level is crucial for allograft survival. Both the drugs have narrow therapeutic range and need to be closely monitored to prevent complications such as nephrotoxicity or rejection. A lot of drugs, food substances and plant ingredients can alter the bioavailability and pharmacokinetics of these two immunosuppressive agents.1,2 This is because first-pass metabolism of both cyclosporine A and sirolimus are largely affected by p-glycoprotein expressed in the gut mucosa and cytochrome CYP3A4 which is mainly present in the liver, although the kidney and gut mucosa also contains this enzyme.

Herbs can interact with the cytochrome P450 system in different ways: by acting as a substrate of one or more of the isoforms of CYP enzymes, or glycoproteins which control their absorption, by acting as an inducer or inhibitor. Among the P450 enzyme, CYP3A family constitutes the most important drug-metabolizing enzymes. They are responsible for metabolizing 50–70% of the drugs.1 The most extensively studied herb is St. John’s Wort. It is a potent inducer of CYP3A4 and p-glycoprotein through activation of pregnane X-receptor.2 Its use in renal transplant patients can cause a low cyclosporine A level resulting in acute allograft rejection.3

On the other hand, herbs or food may inhibit intestinal CYP3A4 enzyme leading to an increase in peak concentration of cyclosporine A level4,5 (Table 1). Grapefruit juice is well known to cause an irreversible inhibition of intestinal CYP3A4 enzyme, leading to loss of enteric CYP3A4 causing a 35% increase in peak concentration of cyclosporine A level.4 Previous reports also showed that Camomile inhibits CYP3A4 and caused increase in cyclosporine level in renal transplant recipients.5 Furthermore, experimental data showed some food like garlic, pepper and cranberry juice or medical herbs like gingko, Siberian Ginseng (Eleutherococcus senticosus), Milk thistle (Silybum marianum), Echinacea (Echinacea Purpura), Licorice and Black cohosh (Cimicifuga racemosa) can inhibit cytochrome CYP3A4. However, their effects in human or cyclosporine level are inconclusive.2

Table 1. Herbs and food substances that interact with cyclosporine A.

	Mechanism	Effect on cyclosporine A level	Evidence in human studies
St. John’s Wort	Induction of CYP3A4 and activity of p- gp	Decrease	Yes
Garlic	constituents increase activity of rat CYP equivalent to CYP3A4 in human	Probably decrease	No
Grapefruit juice  (Pomelo, bitter orange)	Inhibit intestinal CYP3A4	Increase	Yes
Camomile	Inhibit CYP3A4	Increase	Yes
Ginkgo	Inhibit CYP3A4 and ↑ bioavailability of diltiazem in rats	Increase	No
Pepper	Inhibit CYP3A4 and P-gp in vitro	Increase	No
Cranberry juice	Inhibit CYP3A4 mediated nifedipine metabolism in rats	No effect	No

To our knowledge, little has been reported about the rise in cyclosporine and sirolimus levels as a result of interaction with herbal tea. The herbal tea of our patient taken contained three ingredients, oleanolic acid (OA), cimicifuga (CF) and isoliquiritigeni (IS) and was likely the potential cause for the toxic drug levels. The first ingredient, OA, was found to be a potent inhibitor of CYP3A4. Kim et al. tested the inhibitory effect of OA on P450 by incubating OA with human liver microsomes. It showed OA had a competitive inhibition on CYP1A2 and CYP3A4 and the inhibitory potential of OA on CYP3A4-catalyzed midazolam 1-hydroxylation was comparable to some known enzyme inhibitors including cimetidine, diltiazem and clarithromycin.6 In fact, OA is commonly found in the plants worldwide and a popular constituent of herbal medicine. Possible therapeutic effects of OA include antifungal, anti-inflammatory, anti-HIV, diuretic, glucose-lowering and anticancer activities.6

The second ingredient, CF, is also a potent inhibitor of CYP3A4. Tsukamoto et al. demonstrated that commercially available CF had CYP3A4 inhibitory effect comparable to that of ketoconazole.7 CF is a herbal ingredient popular in traditional Chinese medicine. C. heracleifolia, C. dahurica and C. foetida are used as an anti-inflammatory, antipyretic and analgesic remedy. C. racemosa, also known as black cohosh, is used by the Native Americans for the treatment of diarrhea, sore throat and rheumatism. The third ingredient, IS, detected in the herbal tea, is a component in licorice and also a CYP3A4 inhibitor. Tsukamoto et al. isolated 10 compounds from licorice, and looked into their individual inhibitory effect. IS showed some inhibitory effect on CYP3A4 and had an IC 50 of 29 compared with 0.1 of ketoconazole. Glycyrrhizin, the major component of licorice, on the other hand, showed no inhibitory effect on the cytochrome.8 Ursolic acid, which also was found in our patient’s herbal tea, only showed an inhibition on CYP2C19-catalyzed reaction but had no effect on CYP3A4.6 The other ingredients did not have any interaction with CYP3A4.

Based on the above evidences, we believed that the 10-fold increase in cyclosporine A and sirolimus levels was likely due to the recent intake of the herbal tea. It is likely that there was a strong interaction between the immunosuppressive agents and the herbal tea via an inhibitory effect on the CYP3A4 enzyme activity. Nowadays, usage of herbal medication becomes increasingly common. This case alerts us on the consequence of drug–herb interaction and illustrated the toxic drugs levels resulting from taking herbal tea. However, any increase in drug levels may not be noted clinically except through monitoring of blood levels and the nephrotoxic effects of the immunosuppressant drugs are usually indolent.

The acute kidney injury of our patient was likely due to the combined effect of non-steroidal anti-inflammatory drugs (NSAID) and drug–drug interaction between high levels of cyclosporine A and sirolimus. Sirolimus alone seldom causes nephrotoxicity. However, concomitant use with cyclosporine A may enhance the risk of cyclosporine nephrotoxicity. Calcineurin inhibitor and sirolimus are both metabolized by CYP3A4 and compete for extrusion by P-glycoprotein. Sirolimus enhance the cyclosporine toxicity by pushing up the blood and tissue level of both drugs. In addition, sirolimus arrests the cell cycle and may hinder tubular recovery after acute injury.9,10

In conclusion, herbal products may alter the bioavailability and pharmacokinetics of immunosuppressant drugs in many ways and great caution must be taken both by the patients as well as the clinicians. Any increase in drug levels may not be noted clinically. More extensive and in-depth research to understand the properties of these plant-derived foodstuffs would be of great value.

Declaration of interest

The authors have no conflicts of interest to disclose